[Intel-gfx] [PATCH 1/2] drm/i915/gem: Replace reloc chain with terminator on error unwind

[Chris Wilson]

Quoting Pavel Machek (2020-08-19 20:33:26)
> Hi!
> 
> > > > If we hit an error during construction of the reloc chain, we need to
> > > > replace the chain into the next batch with the terminator so that upon
> > > > flushing the relocations so far, we do not execute a hanging batch.
> > > 
> > > Thanks for the patches. I assume this should fix problem from
> > > "5.9-rc1: graphics regression moved from -next to mainline" thread.
> > > 
> > > I have applied them over current -next, and my machine seems to be
> > > working so far (but uptime is less than 30 minutes).
> > > 
> > > If the machine still works tommorow, I'll assume problem is solved.
> > 
> > Aye, best wait until we have to start competing with Chromium for
> > memory... The suspicion is that it was the resource allocation failure
> > path.
> 
> Yep, my machines are low on memory.
> 
> But ... test did not work that well. I have dead X and blinking
> screen. Machine still works reasonably well over ssh, so I guess
> that's an improvement.

> [ 7744.718473] BUG: unable to handle page fault for address: f8c00000
> [ 7744.718484] #PF: supervisor write access in kernel mode
> [ 7744.718487] #PF: error_code(0x0002) - not-present page
> [ 7744.718491] *pdpt = 0000000031b0b001 *pde = 0000000000000000 
> [ 7744.718500] Oops: 0002 [#1] PREEMPT SMP PTI
> [ 7744.718506] CPU: 0 PID: 3004 Comm: Xorg Not tainted 5.9.0-rc1-next-20200819+ #134
> [ 7744.718509] Hardware name: LENOVO 17097HU/17097HU, BIOS 7BETD8WW (2.19 ) 03/31/2011
> [ 7744.718518] EIP: eb_relocate_vma+0xdbf/0xf20

To save me guessing, paste the above location into
	./scripts/decode_stacktrace.sh ./vmlinux . ./drivers/gpu/drm/i915

The f8c00000 is something running off the end of a kmap, but I didn't
spot a path were we would ignore an error and keep on writing.
Nevertheless it must exist.
-Chris